MSTR Cat

Blockchain/Level VI · After bitcoin/6.3 Ethereum and smart contracts/Lesson 5

Stablecoins and the Terra collapse

The peg, the loop, and the spiral

MaterialsA spreadsheet to track daily prices of UST, LUNA, USDT, and DAI from May 2022Access to a block explorer for the Terra blockchain (archived)The SEC complaint against Terraform Labs

Wonder

  1. What mechanism holds a token at $1.00 without holding dollars?
  2. Why does a high yield on deposits accelerate instability rather than dampen it?
  3. How can a collapse in one protocol trigger failures across unrelated lending platforms?

Explore

  1. Stablecoins serve as the base currency for decentralized finance, bridging volatile assets and the real economy. Three primary models exist: fiat-backed, crypto-collateralized, and algorithmic. Fiat-backed tokens like USDT (launched by Tether) and USDC (Circle, September 2018) rely on centralized reserves of dollars or equivalents. Crypto-collateralized tokens like DAI (MakerDAO, December 2017) are backed by over-collateralized loans in volatile assets, subject to liquidation risks if collateral values drop.
  2. Algorithmic stablecoins use economic feedback loops instead of direct backing. TerraUSD (UST), from Do Kwon’s Terraform Labs, maintained its peg through arbitrage: 1 UST could always be swapped for $1 worth of newly minted LUNA. If UST traded below $1, users bought it cheaply to redeem for LUNA at par, shrinking UST supply and theoretically pushing the price back up. This design removed the need for fiat reserves but introduced dependence on market confidence in the redemption mechanism.
  3. The Anchor protocol incentivized holding by paying close to 20% annual yield on UST deposits, subsidized by Terraform Labs and affiliated reserves rather than earned from borrowers alone. This high yield attracted significant capital, creating a large user base dependent on the peg’s stability. The system functioned as long as redemption demand remained manageable relative to LUNA’s market capacity to absorb new supply without crashing its own price.
  4. In May 2022, large-scale withdrawals broke the UST peg. As UST fell below $1, redemptions minted vast quantities of LUNA, which holders dumped on open markets. This flooded LUNA with sell pressure, collapsing its value. With LUNA worthless, the redemption mechanism lost its backing, causing UST to free-fall. The reflexive loop accelerated: a falling peg triggered more minting and dumping, erasing tens of billions in market value within days.
  5. Contagion spread rapidly. Three Arrows Capital, a hedge fund heavily leveraged into Terra assets, collapsed. The lender Celsius froze withdrawals in June 2022 amid losses and a run, and Voyager failed after Three Arrows defaulted on a large loan. The failure demonstrated that algorithmic designs lack the external anchor of fiat reserves, making them vulnerable to reflexivity when market confidence evaporates.
  6. Do Kwon was arrested in Montenegro in March 2023 and extradited to the United States in December 2024. In August 2025 he pleaded guilty in New York to fraud charges, and in December 2025 he was sentenced to 15 years in prison. The Terra collapse remains a defining case study in DeFi risk: code alone cannot guarantee stability without underlying asset backing or robust economic buffers against reflexive selling pressure.

Key facts

  • USDC launched in September 2018; DAI launched in December 2017.
  • TerraUSD (UST) aimed to hold its peg because 1 UST could always be swapped for $1 worth of newly minted LUNA.
  • The Anchor protocol paid approximately 20% yield on UST deposits.
  • In May 2022, UST depegged, triggering a death spiral that erased billions in value.
  • Three Arrows Capital, Celsius and Voyager failed in the turmoil that followed Terra’s collapse.
  • Do Kwon was extradited to the US in December 2024.

Exercise nine, six, three, one

9CollectList 9 stablecoins or pegged tokens. Note their backing mechanism: fiat reserves, crypto collateral, algorithmic minting/burning, or commodity.
6ClusterGroup the 9 into 6 by risk profile: fully reserved, over-collateralized, under-collateralized, purely algorithmic, hybrid, or opaque/centralized control.
3DistillReduce the 6 to 3 distinct failure modes observed in historical collapses (e.g., reserve shortfall, liquidation cascade, reflexivity).
1CrystallizeWrite 1 sentence on what a peg promises and which specific asset or rule keeps that promise honest.

Fact check

Dates, figures and claims checked against the published sources listed below.

Open in the interactive academy →

PT ES ZH JA KO RU DE FR MS